- 25 Jan, 2020 2 commits
-
-
Tyrel Datwyler authored
Commit e5afdf9d ("powerpc/vfio_spapr_tce: Add reference counting to iommu_table") missed an iommu_table allocation in the pseries vio code. The iommu_table is allocated with kzalloc and as a result the associated kref gets a value of zero. This has the side effect that during a DLPAR remove of the associated virtual IOA the iommu_tce_table_put() triggers a use-after-free underflow warning. Call Trace: [c0000002879e39f0] [c00000000071ecb4] refcount_warn_saturate+0x184/0x190 (unreliable) [c0000002879e3a50] [c0000000000500ac] iommu_tce_table_put+0x9c/0xb0 [c0000002879e3a70] [c0000000000f54e4] vio_dev_release+0x34/0x70 [c0000002879e3aa0] [c00000000087cfa4] device_release+0x54/0xf0 [c0000002879e3b10] [c000000000d64c84] kobject_cleanup+0xa4/0x240 [c0000002879e3b90] [c00000000087d358] put_device+0x28/0x40 [c0000002879e3bb0] [c0000000007a328c] dlpar_remove_slot+0x15c/0x250 [c0000002879e3c50] [c0000000007a348c] remove_slot_store+0xac/0xf0 [c0000002879e3cd0] [c000000000d64220] kobj_attr_store+0x30/0x60 [c0000002879e3cf0] [c0000000004ff13c] sysfs_kf_write+0x6c/0xa0 [c0000002879e3d10] [c0000000004fde4c] kernfs_fop_write+0x18c/0x260 [c0000002879e3d60] [c000000000410f3c] __vfs_write+0x3c/0x70 [c0000002879e3d80] [c000000000415408] vfs_write+0xc8/0x250 [c0000002879e3dd0] [c0000000004157dc] ksys_write+0x7c/0x120 [c0000002879e3e20] [c00000000000b278] system_call+0x5c/0x68 Further, since the refcount was always zero the iommu_tce_table_put() fails to call the iommu_table release function resulting in a leak. Fix this issue be initilizing the iommu_table kref immediately after allocation. Fixes: e5afdf9d ("powerpc/vfio_spapr_tce: Add reference counting to iommu_table") Signed-off-by:
Tyrel Datwyler <tyreld@linux.ibm.com> Reviewed-by:
Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by:
Michael Ellerman <mpe@ellerman.id.au> Link: https://lore.kernel.org/r/1579558202-26052-1-git-send-email-tyreld@linux.ibm.com
-
Aneesh Kumar K.V authored
Setting ND_REGION_PAGEMAP flag implies namespace mode defaults to fsdax mode. This also means kernel ends up creating struct page backing for these namspace ranges. With large namespaces that is not the right thing to do. We should let the user select the mode he/she wants the namespace to be created with. Hence disable ND_REGION_PAGEMAP for papr_scm regions. We still keep the flag for of_pmem because it supports only small persistent memory regions. This is similar to what is done for x86 with commit commit: 004f1afb ("libnvdimm, pmem: direct map legacy pmem by default") Signed-off-by:
Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com> Signed-off-by:
-
- 23 Jan, 2020 38 commits
-
-
Greg Kurz authored
As reported by ./scripts/checkpatch.pl --strict: CHECK: extern prototypes should be avoided in .h files Signed-off-by:
Greg Kurz <groug@kaod.org> Signed-off-by:
-
Greg Kurz authored
Signed-off-by:
-
Krzysztof Kozlowski authored
Adjust indentation from spaces to tab (+optional two spaces) as in coding style with command like: $ sed -e 's/^ /\t/' -i */Kconfig Signed-off-by:
Krzysztof Kozlowski <krzk@kernel.org> Signed-off-by:
-
Laurentiu Tudor authored
Michael Ellerman made a call for volunteers from NXP to maintain this driver and I offered myself. Signed-off-by:
Laurentiu Tudor <laurentiu.tudor@nxp.com> Acked-by:
Timur Tabi <timur@kernel.org> Signed-off-by:
-
Oliver O'Halloran authored
This is only used in pci-ioda.c so move it there and rename it to match. Signed-off-by:
Oliver O'Halloran <oohall@gmail.com> Reviewed-by:
-
Oliver O'Halloran authored
pnv_pci_dma_dev_setup() does nothing but call the phb->dma_dev_setup() callback, if one exists. That callback is only set for normal PCIe PHBs so we can remove the layer of indirection and use the ioda version in the pci_controller_ops. Signed-off-by:
-
Oliver O'Halloran authored
An ioda_pe for each VF is allocated in pnv_pci_sriov_enable() before the pci_dev for the VF is created. We need to set the pe->pdev pointer at some point after the pci_dev is created. Currently we do that in: pcibios_bus_add_device() pnv_pci_dma_dev_setup() (via phb->ops.dma_dev_setup) /* fixup is done here */ pnv_pci_ioda_dma_dev_setup() (via pnv_phb->dma_dev_setup) The fixup needs to be done before setting up DMA for for the VF's PE, but there's no real reason to delay it until this point. Move the fixup into pnv_pci_ioda_fixup_iov() so the ordering is: pcibios_add_device() pnv_pci_ioda_fixup_iov() (via ppc_md.pcibios_fixup_sriov) pcibios_bus_add_device() ... This isn't strictly required, but it's a slightly more logical place to do the fixup and it simplifies pnv_pci_dma_dev_setup(). Signed-off-by:
-
Oliver O'Halloran authored
The pnv_pci_dma_dev_setup() only does something when: 1) There PHB contains VFs, or 2) The PHB defines a dma_dev_setup() callback in the pnv_phb structure. Neither is true for NPU PHBs so there's no reason to set the callback. Reviewed-by:
-
Oliver O'Halloran authored
pcibios_bus_add_device() is the only caller of pcibios_setup_device(). Fold them together since there's no real reason to keep them separate. Signed-off-by:
-
Oliver O'Halloran authored
OPAL provides several different kinds of reboot for the kernel to use, namely forcing a full reboot, platform error reboot and MPIPL. Right now triggering the alternative resets requires some ad-hoc method such as triggering a kernel crash and hoping the stars align. It's sometimes handy to be able to trigger one of these resets directly, so add a way to do that. Signed-off-by:
-
Oliver O'Halloran authored
On PowerNV a few different kinds of reboot are supported. We'd like to be able to exercise these from xmon so allow 'zr' to take an argument, and pass that to the ppc_md.restart() function. Signed-off-by:
-
Oliver O'Halloran authored
Long before we had a generic way for firmware to export memory ranges of interest we added a special case for the skiboot symbol map. The code is pretty much identical to the generic export so re-use the code. Signed-off-by:
-
Oliver O'Halloran authored
Originally we only had a handful of exported memory ranges, but we'd to export the per-core trace buffers. This results in a lot of files in the exports directory which is a but unfortunate. We can clean things up a bit by turning subnodes into subdirectories of the exports directory. Signed-off-by:
-
Oliver O'Halloran authored
Many drivers don't check for errors when they get a 0xFFs response from an MMIO load. As a result after an EEH event occurs a driver can get stuck in a polling loop unless it some kind of internal timeout logic. Currently EEH tries to detect and report stuck drivers by dumping a stack trace after eeh_dev_check_failure() is called EEH_MAX_FAILS times on an already frozen PE. The value of EEH_MAX_FAILS was chosen so that a dump would occur every few seconds if the driver was spinning in a loop. This results in a lot of spurious stack traces in the kernel log. Fix this by limiting it to printing one stack trace for each PE freeze. If the driver is truely stuck the kernel's hung task detector is better suited to reporting the probelm anyway. Signed-off-by:
Sam Bobroff <sbobroff@linux.ibm.com> Tested-by:
-
Oliver O'Halloran authored
Add a debugfs entry to dump the state of the active IODA PEs. The IODA PE state reflects how the PHB's internal concept of a PE is configured. This is separate to the EEH PE state and is managed power the PowerNV PCI backend rather than the EEH core. Signed-off-by:
-
Oliver O'Halloran authored
Make the dump trigger off any input rather than just '1'. This allows you to write "echo 1> dump_diag_data" and it'll do what you want rather than erroring out pointlessly. Signed-off-by:
-
Oliver O'Halloran authored
Use the pnv_phb structure as the private data pointer for the debugfs files. This lets us delete some code and an open-coded use of hose->private_data. Signed-off-by:
-
Oliver O'Halloran authored
These functions can only be used on a SR-IOV capable physical function and they're only called in pcibios_sriov_enable / disable. Make them emit a warning in the future if they're used incorrectly and remove the dead code that checks if the device is a VF. Signed-off-by:
-
Oliver O'Halloran authored
The powerpc PCI code requires that a pci_dn structure exists for all devices in the system. This is fine for real devices since at boot a pci_dn is created for each PCI device in the DT and it's fine for hotplugged devices since the hotplug slot driver will manage the pci_dn's devices in hotplug slots. For SR-IOV, we need the platform / pcibios to manage the pci_dn for virtual functions since firmware is unaware of VFs, and they aren't "hot plugged" in the traditional sense. Management of the pci_dn is handled by the, poorly named, functions: add_pci_dev_data() and remove_pci_dev_data(). The entire body of these functions is #ifdef`ed around CONFIG_PCI_IOV and they cannot be used in any other context, so make them only available when CONFIG_PCI_IOV is selected, and rename them to reflect their actual usage rather than having them masquerade as generic code. Signed-off-by:
-
Oliver O'Halloran authored
When disabling virtual functions on an SR-IOV adapter we currently do not correctly remove the EEH state for the now-dead virtual functions. When removing the pci_dn that was created for the VF when SR-IOV was enabled we free the corresponding eeh_dev without removing it from the child device list of the eeh_pe that contained it. This can result in crashes due to the use-after-free. Signed-off-by:
-
Oliver O'Halloran authored
The eeh_sysfs_remove_device() function is supposed to clear the EEH_DEV_SYSFS flag since it indicates the EEH sysfs entries have been added for a pci_dev. When the sysfs files are removed eeh_remove_device() the eeh_dev and the pci_dev have already been de-associated. This then causes the pci_dev_to_eeh_dev() call in eeh_sysfs_remove_device() to return NULL so the flag can't be cleared from the still-live eeh_dev. This problem is worked around in the caller by clearing the flag manually. However, this behaviour doesn't make a whole lot of sense, so this patch fixes it by: a) Re-ordering eeh_remove_device() so that eeh_sysfs_remove_device() is called before de-associating the pci_dev and eeh_dev. b) Making eeh_sysfs_remove_device() emit a warning if there's no corresponding eeh_dev for a pci_dev. The paths where the sysfs files are only reachable if EEH was setup for the device for the device in the first place so hitting this warning indicates a programming error. Signed-off-by:
-
Oliver O'Halloran authored
In eeh_notify_resume_show() the pci_dn for the device is looked up once in the declaration block and then once after checking for a NULL eeh_dev. Remove the second lookup since it's pointless. Signed-off-by:
-
Oliver O'Halloran authored
There are several EEH sysfs properties that only exists when the "ibm,is-open-sriov-pf" property appears in the device tree node of the PCI device. This used on pseries to indicate to the guest that the hypervisor allows the guest to configure the SR-IOV capability. Doing this requires some handshaking between the guest, hypervisor and userspace when a VF is EEH frozen which is why these properties exist. This is all dead code on non-pseries platforms so wrap it in an #ifdef CONFIG_PPC_PSERIES to make the dependency clearer. Signed-off-by:
-
Oliver O'Halloran authored
The EEH_ATTR_SHOW() helper is used to display fields from struct eeh_dev not struct pci_dn. Signed-off-by:
-
Oliver O'Halloran authored
At the point where we start inserting ranges into the EEH address cache the binding between pci_dev and eeh_dev has already been set up. Instead of consulting the pci_dn tree we can retrieve the eeh_dev directly using pci_dev_to_eeh_dev(). Signed-off-by:
-
Frederic Barrat authored
The PCI hotplug framework is used to update the devices when a new image is written to the FPGA. Reviewed-by:
Alastair D'Silva <alastair@d-silva.org> Reviewed-by:
Andrew Donnellan <ajd@linux.ibm.com> Signed-off-by:
Frederic Barrat <fbarrat@linux.ibm.com> Signed-off-by:
-
Frederic Barrat authored
An opencapi slot doesn't have an associated bridge device. It's not needed for operation, but any warning is displayed through pci_warn() which uses the pci_dev struct of the assocated bridge device. So wrap those warning so that a different trace mechanism can be used if it's an opencapi slot. Reviewed-by:
-
Frederic Barrat authored
The driver only allows to disable a slot in the POPULATED state. However, if an error occurs while enabling the slot, say because the link couldn't be trained, then the POPULATED state may not be reached, yet the power state of the slot is on. So allow to disable a slot in the REGISTERED state. Removing the devices will do nothing since it's not populated, and we'll set the power state of the slot back to off. Reviewed-by:
-
Frederic Barrat authored
Add the opencapi PHBs to the list of PHBs being scanned to look for slots. Signed-off-by:
-
Frederic Barrat authored
When changing the slot state, if opal hits an error and tells as such in the asynchronous reply, the warning "Wrong msg" is logged, which is rather confusing. Instead we can reuse the better message which is already used when we couldn't submit the asynchronous opal request initially. Reviewed-by:
-
Frederic Barrat authored
On powernv, when removing a device through hotplug, the following warning is logged: Invalid refcount <.> on <...> It may be incorrect, the refcount may be set to a higher value than 1 and be valid. of_detach_node() can drop more than one reference. As it doesn't seem trivial to assert the correct value, let's remove the warning. Reviewed-by: -
Frederic Barrat authored
Unlike real PCI slots, opencapi slots are directly associated to the (virtual) opencapi PHB, there's no intermediate bridge. So when looking for a slot ID, we must start the search from the device node itself and not its parent. Also, the slot ID is not attached to a specific bdfn, so let's build it from the PHB ID, like skiboot. Signed-off-by:
-
Frederic Barrat authored
With hotplug, an opencapi device can now go away. It needs to be released, mostly to clean up its PE state. We were previously not defining any device callback. We can reuse the standard PCI release callback, it does a bit too much for an opencapi device, but it's harmless, and only needs minor tuning. Also separate the undo of the PELT-V code in a separate function, it is not needed for NPU devices and it improves a bit the readability of the code. Signed-off-by:
-
Frederic Barrat authored
The PE for an opencapi device was set as part of a late PHB fixup operation, when creating the PHB. To use the PCI hotplug framework, this is not going to work, as the PHB stays the same, it's only the devices underneath which are updated. For regular PCI devices, it is done as part of the reconfiguration of the bridge, but for opencapi PHBs, we don't have an intermediate bridge. So let's define the PE when the device is enabled. PEs are meaningless for opencapi, the NPU doesn't define them and opal is not doing anything with them. Reviewed-by:
-
Frederic Barrat authored
Protect the PHB's list of PE. Probably not needed as long as it was populated during PHB creation, but it feels right and will become required once we can add/remove opencapi devices on hotplug. Reviewed-by:
-
Frederic Barrat authored
The pci_dn structure used to store a pointer to the struct pci_dev, so taking a reference on the device was required. However, the pci_dev pointer was later removed from the pci_dn structure, but the reference was kept for the npu device. See commit 902bdc57 ("powerpc/powernv/idoa: Remove unnecessary pcidev from pci_dn"). We don't need to take a reference on the device when assigning the PE as the struct pnv_ioda_pe is cleaned up at the same time as the (physical) device is released. Doing so prevents the device from being released, which is a problem for opencapi devices, since we want to be able to remove them through PCI hotplug. Now the ugly part: nvlink npu devices are not meant to be released. Because of the above, we've always leaked a reference and simply removing it now is dangerous and would likely require more work. There's currently no release device callback for nvlink devices for example. So to be safe, this patch leaks a reference on the npu device, but only for nvlink and not opencapi. Signed-off-by:
-
Christophe Leroy authored
Various optimisations by inverting branches and removing redundant instructions. Signed-off-by:
Christophe Leroy <christophe.leroy@c-s.fr> Signed-off-by:
-
Christophe Leroy authored
clock_getres returns hrtimer_res for all clocks but coarse ones for which it returns KTIME_LOW_RES. return EINVAL for unknown clocks. Signed-off-by:
-
